codingplans.cc

Compare providers and see every monthly plan tier at a glance.

Showing 79 providers

Direct Providers

12 providers
Brand Lite / Entry Pro / Standard Plus / Max Scale / Ultra Privacy Artificial Analysis Coding Index Action
OpenAI / Codex
HQ: San Francisco, CA, US
Models: GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.6 Luna, GPT-5.5, GPT-5.4, GPT-5.4 mini
Access: Codex CLI, Codex IDE extension, Codex web, Codex iOS, OpenClaw, OpenAI-compatible API
Free$0/mo
Limited Codex access; GPT-5.6 Terra
Go$8/mo
Lightweight Codex; GPT-5.6 Terra
Plus$20/mo
Expanded Codex; GPT-5.6 Sol/Terra/Luna
Pro 5×$100/mo
5× Plus limits; GPT-5.6 Sol Pro; Ultra mode
Pro 20×$200/mo
20× Plus limits; GPT-5.6 Sol Pro; Ultra mode
 Standard
TrainingMay use
Prompt retentionVariable
PrivacyData controls
78.3 Artificial Analysis Coding Index
GPT-5.6
GPT-5.6 Sol (xhigh)
GPT-5.6 Sol is the Codex flagship; Sol (xhigh) is the highest-scoring GPT-5.6 variant. GPT-5.5 remains available.
Estimated 61.8 tok/s
 View
Anthropic / Claude
HQ: San Francisco, CA, US
Models: Claude Opus 5, Claude Sonnet 5, Claude Fable 5, Claude Opus 4.8, Claude Sonnet 4.6, Claude Haiku 4.5
Access: Claude Code, Claude Cowork, Claude Design, Claude Science
Note: As of Jul 22 2026, Fable 5 is available only to Max 5× and Max 20× subscribers at 50% of limits (not Pro). Model traini…
Pro$20/mo
Higher use; Opus 4.8, Sonnet 4.6, Haiku 4.5
Max 5×$100/mo
5× Pro limits; includes Fable 5
Max 20×$200/mo
20× Pro limits; includes Fable 5
 — Partial
TrainingPartial
Prompt retentionVariable
PrivacyTraining
78 Artificial Analysis Coding Index
Claude Opus 5
Claude Opus 5 (Adaptive Reasoning, Max Effort)
Estimated 52.6 tok/s
 View
Kimi / Moonshot
HQ: Beijing, China / Singapore
Models: Kimi K3, Kimi K2.6 (free in chat), kimi-for-coding (1M ctx)
Access: Kimi Code CLI, Kimi Code VS Code, Claude Code, Roo Code, OpenCode, OpenAI-compatible
Note: Adagio (free) is chat-only; Kimi Code is a paid-only benefit. K3 supports up to 1M context tokens.
Moderato$19/mo
60 agent credits; Kimi Code 1×; 2 con tasks
Allegretto$39/mo
150 agent credits; Kimi Code 5×; Kimi Claw
Allegro$99/mo
360 agent credits; Kimi Code 15×; 4 con tasks
Vivace$199/mo
720 agent credits; Kimi Code 30×; 4 con tasks; max Swarm
 Standard
TrainingMay use
Prompt retentionVariable
PrivacyAPI privacy
76.2 Artificial Analysis Coding Index
Kimi K3
Kimi K3 (max)
Kimi K3 is the current flagship; K2.6 remains free in chat and does not consume credits.
Estimated 37.8 tok/s
 View
Z.AI GLM
HQ: Beijing, China
Models: GLM-5.3, GLM-5.2, GLM-5-Turbo
Access: All major coding tools, OpenAI-compatible, Anthropic-compatible
Note: GLM-5.3 released; usage policy is a 5-hour limit plus weekly credit quota. Quarterly billing −20%; yearly billing −30%.
Lite$18/mo
10K credits/wk; rolling access to latest flagship models
Pro$80/mo
6× Lite usage; priority flagship access; curated MCP tools; faster generation
Max$168/mo
14× Lite usage; first flagship access; dedicated peak resources
 — Partial
TrainingPartial
Prompt retentionVariable
PrivacyTerms
74.8 Artificial Analysis Coding Index
GLM-5.3
GLM-5.3 (max)
Estimated 80.4 tok/s
 View
xAI / Grok Build
HQ: Palo Alto, CA, US
Models: Grok Build Beta, Grok 4.5, Grok 4 Heavy, Grok Code Fast 1
Access: Grok Build CLI, Plan mode, headless mode (-p), ACP, MCP, AGENTS.md, Imagine
Note: Grok Build beta is available to SuperGrok and SuperGrok Heavy subscribers; previously Heavy-only early access.
SuperGrok$30/mo
Grok Build beta access; Grok 4.5
SuperGrok Heavy$300/mo
Highest limits; Grok 4 Heavy; $99/mo promo for first 3 months
 — — Partial
TrainingPartial
Prompt retentionLimited
Grok BuildPrivacyConsumer FAQTerms
72.4 Artificial Analysis Coding Index
Grok 4.5
Grok 4.5 (high)
Grok 4.5 is the current xAI coding flagship (launched Jul 8 2026); Grok Build uses grok-build-0.1.
Estimated 51 tok/s
 View
Google Gemini / Code Assist
HQ: Mountain View, CA, US
Models: Gemini 3.6 Flash, Gemini 3.1 Pro, Gemini 3.1 Deep Think, Gemini Spark
Access: Antigravity CLI, Antigravity, Jules, Gemini app
Note: Gemini CLI shut down Jun 18 2026; consumer tiers migrated to Antigravity CLI. AI Ultra split into 5× ($99.99) and 20× (…
Google AI Plus$4.99/mo
2× Free usage; 400GB storage; Gemini 3.6 Flash
Google AI Pro$19.99/mo
4× Free usage; Gemini 3.1 Pro; Jules; Antigravity entry limits
Google AI Ultra 5×$99.99/mo
5× Pro limits; Deep Think; Gemini Spark; Antigravity Ultra 5×
Google AI Ultra 20×$199.99/mo
20× Pro limits; Deep Think; Gemini Spark; Antigravity Ultra 20×
 Partial
TrainingPartial
Prompt retentionVariable
PrivacyGemini privacy
69.2 Artificial Analysis Coding Index
Gemini 3.6 Flash
Gemini 3.6 Flash (high)
Gemini 3.6 Flash is available on Google AI plans and scores higher than Gemini 3.1 Pro Preview; 3.1 Pro and Deep Think remain available on higher tiers.
Estimated 201.2 tok/s
 View
BytePlus ModelArk
HQ: Singapore / ByteDance; processing in Malaysia, Indonesia and/or EU/EEA
Models: DeepSeek-V4-Pro, DeepSeek-V4-Flash, Dola-Seed-2.0-pro, Dola-Seed-2.0-Code, GLM-5.2, GLM-4.7, Kimi-K2.5, GPT-OSS
Access: ArkClaw, Cursor, Roo Code, OpenAI-compatible tools
Lite$10/mo
3× Claude Pro usage
Pro$50/mo
5× Lite usage; bundles ArkClaw
TeamCustom
Per-seat team plan; central billing and admin; contact sales
 — Partial
TrainingPartial
Prompt retentionVariable
Data processingPrivacy
68.8 Artificial Analysis Coding Index
DeepSeek V4 Pro
DeepSeek V4 Pro 0813 (Reasoning, Max Effort)
Estimated 79.1 tok/s
 View
Xiaomi MiMo
HQ: Beijing, China
Models: MiMo-V2.5-Pro (1M ctx; 2× credits), MiMo-V2.5-Pro-UltraSpeed (1T; 1000 tok/s), MiMo-V2.5 (1M ctx; 1× credits), MiMo-V2.5-TTS, MiMo-V2.5-TTS-VoiceDesign, MiMo-V2.5-TTS-VoiceClone, MiMo-V2.5-ASR
Access: MiMo Claw, OpenClaw, Claude Code, OpenCode, KiloCode, OpenAI-compatible, Anthropic-compatible
Note: MiMo-V2.5-Pro-UltraSpeed is API-only (not in Token Plan). MiMo Claw limited-time offer ¥14.9/mo.
Lite$6/mo
4.1B credits (intro: $4.20/mo)
Standard$16/mo
11B credits (intro: $11.20/mo)
Pro$50/mo
38B credits (intro: $35/mo)
Max$100/mo
82B credits (intro: $70/mo)
 Standard
TrainingMay use
Prompt retentionStored
PrivacyToken PlanMiMo
60.2 Artificial Analysis Coding Index
MiMo-V2.5-Pro
MiMo-V2.5-Pro
Estimated 58.3 tok/s
 View
StreamLake KwaiKAT Coding Plan
HQ: Singapore (StreamLake Pte. Ltd.)
Models: KAT-Coder-Pro V2.5 (256K ctx; 80K max output), KAT-Coder-Pro V2, KAT-Coder-Pro V1
Access: Claude Code, Cline, Kilo, OpenCode, OpenClaw, OpenAI-compatible API, Claude-compatible proxy
Note: Fixed 5-hour prompt windows; one frontend Prompt ≈ 15-20 backend model calls. KAT-Coder-Pro V2.5 launched Jul 10 2026.
Mini$5/mo
40 prompts/5h (~600 req); KAT-Coder-Pro V2.5; 60 RPM / 2M TPM
Starter$10/mo
100 prompts/5h (~1,500 req); optimized for Claude Code and major IDEs
Pro$20/mo
300 prompts/5h (~4,500 req); 3x Starter usage
Max$50/mo
1,000 prompts/5h (~15,000 req); repository-level coding use
 Standard
TrainingMay use
Prompt retentionStored
Coding PlanKAT-CoderPrivacyTerms
59.5 Artificial Analysis Coding Index
KAT-Coder-Pro V2
KAT Coder Pro V2
KAT-Coder-Pro V2.5 launched Jul 10 2026 but is not yet benchmarked by Artificial Analysis; Coding Index shown for V2.
Estimated 109 tok/s
 View
MiniMax Token Plan
HQ: Shanghai, China / Singapore
Models: MiniMax M3, MiniMax M2.7, image generation, speech synthesis, music generation
Access: All major coding tools, OpenAI-compatible
Plus$20/mo
~1.7B tok/mo (M3); 3-4 con agents; text, image, speech, music
Max$50/mo
~5.1B tok/mo (M3); 4-5 con agents; text, image, speech, music; 3 video clips/d
Ultra$120/mo
~9.8B tok/mo (M3); 6-7 con agents; text, image, speech, music; 5 video clips/d
 — Partial
TrainingPartial
Prompt retentionStored
API privacyPrivacy
58.6 Artificial Analysis Coding Index
MiniMax-M3
MiniMax-M3
Estimated 104.4 tok/s
 View
Mistral AI
HQ: Paris, France
Models: Mistral Medium 3.5, Mistral Small 4, Mistral Large 3, Devstral 2, Devstral Small 2, Codestral, Magistral Medium, Magistral Small, Ministral 3B/8B/14B, Voxtral, Mistral OCR 4
Access: Mistral Vibe, Vibe CLI, Vibe for IDE, Le Chat, OpenAI-compatible
Free$0/mo
Limited Vibe access; latest models; limited coding sessions
Pro$14.99/mo
Full Vibe access; all-day coding; more messages/search/image gen
Team$24.99/mo
Per user; team workspace; 30GB storage/user; domain verification; data export
 — Partial
TrainingPartial
Prompt retentionVariable
PrivacyTraining
46.9 Artificial Analysis Coding Index
Mistral Medium 3.5
Mistral Medium 3.5
Estimated 146.5 tok/s
 View
StepFun Step Plan
HQ: Shanghai, China
Models: step-3.7-flash (256K ctx; multimodal reasoning), step-3.5-flash-2603, step-3.5-flash
Access: OpenClaw, Claude Code, Trae, Cursor, Cline, Goose, Kilo Code, Roo Code, Step Plan API
Note: One prompt is roughly 15-20 model calls; Studio credits add +40% quota.
Flash Mini$6.99/mo
100 prompts/5h; 400/wk (~1.5K req/5h)
Flash Plus$9.99/mo
400 prompts/5h; 1.6K/wk (~6K req/5h)
Flash Pro$29/mo
1.5K prompts/5h; 6K/wk (~22.5K req/5h)
Flash Max$99/mo
5K prompts/5h; 20K/wk (~75K req/5h)
 Standard
TrainingMay use
Prompt retentionVariable
PrivacyStep Plan
39.6 Artificial Analysis Coding Index
Step 3.7 Flash
Step 3.7 Flash
Estimated 120.1 tok/s
 View

Aggregators & Routers

24 providers
Brand Lite / Entry Pro / Standard Plus / Max Scale / Ultra Privacy Artificial Analysis Coding Index Action
Apertis Coding Plan
HQ: Wyoming, US
Models: Claude Opus 4.8, Claude Opus 4.7, Claude Sonnet 5, Claude Sonnet 4.6, GPT-5.6, GPT-5.5, Gemini 3.6 Flash, Gemini 3.1 Pro Preview, DeepSeek V4 Pro, GLM-5.2, Kimi K3, Kimi K2.6, MiniMax M3, Grok 4.5, Qwen3.7
Access: Claude Code, Cursor, OpenClaw, OpenCode, Cline, Roo Code, Kilo Code, Aider, Goose, Crush, OpenAI-compatible API, Anthropic-compatible API
Note: Subscription quota is consumed per request by model multiplier. Cache hits consume no quota. PAYG fallback is optional.
Lite$12/mo
600 quota/mo; fixed quota per request by model multiplier; free cache hits
Pro$25/mo
900 quota/mo; fixed quota per request by model multiplier; web search MCP
Plus$60/mo
1.5K quota/mo; lower per-request model rates; priority support
Max$200/mo
3K quota/mo; guaranteed peak-hour performance; early access
 Partial
TrainingPartial
Prompt retentionLimited
SubscribeDocsPrivacyTrustTerms
78.3 Artificial Analysis Coding Index
GPT-5.6
GPT-5.6 Sol (xhigh)
 View
LLM Gateway DevPass
HQ: Delaware, US
Models: Claude Fable 5, Claude Opus 4.8, Claude Sonnet 5, GPT-5.6, GPT-5.5 Pro, GPT-5.5, Claude Opus 4.7, Gemini 3.6 Flash, Gemini 3.1 Pro Preview, GLM-5.2, Kimi K2.7, Kimi K2.6, DeepSeek V4 Pro, Grok 4.5, MiniMax M3, 200+ model catalog
Access: Claude Code, OpenCode, SoulForge, Cline, OpenAI-compatible tools
Note: DevPass includes 3× provider-rate usage allowance; SoulForge can roughly double effective usage on coding-agent work.
Lite$29/mo
$87 provider-rate value
Pro$79/mo
$237 provider-rate value
Max$179/mo
$537 provider-rate value
 — Partial
TrainingMay use
Prompt retentionPartial
PrivacyRetention
78.3 Artificial Analysis Coding Index
GPT-5.6
GPT-5.6 Sol (xhigh)
 View
Routera
HQ: Deva, Hunedoara, Romania
Models: GPT 5.6 Sol, GPT 5.6 Terra, GPT 5.6 Luna, GPT 5.5, Claude Opus 4.8, Claude Opus 4.7, Claude Opus 4.6, Claude Sonnet 4.6, Claude Fable 5, Grok 4.5, Kimi 2.6, GLM 5.1, MiniMax 2.7, Gemma 4 31B, Qwen3.5 27B
Access: OpenAI-compatible API, Chat Completions, Responses API, streaming, code assistance, agentic systems, search/RAG workflows
GPT Plan$90/mo
500M GPT tokens/mo; GPT 5.6 Sol/Terra/Luna + GPT 5.5; OpenAI-compatible API
Claude Plan$120/mo
500M Claude tokens/mo; Opus 4.8/4.7/4.6, Sonnet 4.6; best for complex reasoning
GPT + Claude$200/mo
1B shared tokens/mo across GPT and Claude; all subscription models; best value for mixed workloads
Pay as you go$30
100M all-model tokens; $3 per 10M tokens; never expire or reset
 ZDR
TrainingNo
Prompt retentionZero
PricingModelsCode assistanceAgentic systemsPrivacyTerms
78.3 Artificial Analysis Coding Index
GPT-5.6 Sol
GPT-5.6 Sol (xhigh)
 View
Nous Portal
HQ: US (Nous Research)
Models: anthropic/claude-opus-5, anthropic/claude-sonnet-5, anthropic/claude-fable-5, openai/gpt-5.6-sol, openai/gpt-5.6-terra, openai/gpt-5.6-luna, openai/gpt-5.5, google/gemini-3.6-flash, google/gemini-3.5-flash, google/gemini-3.1-pro-preview, x-ai/grok-4.5, qwen/qwen3.8-max, qwen/qwen3.7-max, moonshotai/kimi-k3, moonshotai/kimi-k2.7-code, z-ai/glm-5.2 +6
Access: Hermes Agent, Nous Chat, OpenAI-compatible
Free$0/mo
Limited access
Plus$20/mo
$22 credits; $10 rollover cap
Super$100/mo
$110 credits; $50 rollover cap
Ultra$200/mo
$220 credits; $100 rollover cap
 Standard
TrainingMay use
Prompt retentionVariable
Privacy
78 Artificial Analysis Coding Index
Claude Opus 5
Claude Opus 5 (Adaptive Reasoning, Max Effort)
 View
ZenMux
HQ: Singapore
Models: GPT-5.6, GPT-5.5 Pro, GPT-5.5, GPT-5.4 Pro, GPT-5.4, GPT-5.3-Codex, Claude Opus 5, Claude Fable 5, Claude Opus 4.8, Claude Opus 4.7, Claude Sonnet 5, Claude Sonnet 4.6, Gemini 3.6 Flash, Gemini 3.1 Pro Preview, Qwen3.8 Max, Qwen3.7 Max +11
Access: Studio Chat, API Request, Codex, OpenCode, Cline, Cursor, OpenAI-compatible API, Anthropic-compatible API
Free$0/mo
$0.16/5h PAYG value; $5/mo PAYG value; Studio Chat only
Pro$20/mo
$1.64/5h PAYG value; $30/mo PAYG value; Studio Chat + API
Max$100/mo
$9.85/5h PAYG value; $180/mo PAYG value; Pro/video models
Ultra$200/mo
$26.27/5h PAYG value; $480/mo PAYG value; all models
 Standard
TrainingMay use
Prompt retentionVariable
PrivacyTermsSubscription docsModels
78 Artificial Analysis Coding Index
Claude Opus 5
Claude Opus 5 (Adaptive Reasoning, Max Effort)
 View
Chutes
HQ: Canada (Chutes Global Corp.)
Models: moonshotai/Kimi-K3-TEE, moonshotai/Kimi-K2.6-TEE, zai-org/GLM-5.2-TEE, zai-org/GLM-5.1-TEE, deepseek-ai/DeepSeek-V4-Flash-0731-TEE, deepseek-ai/DeepSeek-V3.2-TEE, Qwen/Qwen3.6-27B-TEE, Qwen/Qwen3-235B-A22B-Thinking-2507-TEE
Access: API-compatible coding tools
Plus$10/mo
5× PAYG; frontier included
Pro$20/mo
5× PAYG; 10% discount
 — — ZDR
TrainingNo
Prompt retentionZero
Privacy
76.2 Artificial Analysis Coding Index
Kimi K3
Kimi K3 (max)
 View
Neuralwatt Cloud
HQ: Seattle, WA, US
Models: GLM-5.2, GLM-5.2 Fast, Kimi K3, Kimi K3 Fast, Kimi K2.7 Code, Kimi K2.7 Code Fast, Qwen3.6 35B, Qwen3.6 35B Fast, DeepSeek V4 Flash
Access: OpenAI-compatible API, Claude Code, OpenCode
Basic$20/mo
2.35 kWh included; overage at $8.50/kWh
Standard$50/mo
6.25 kWh included; priority access during high demand; overage at $8.00/kWh
Pro$100/mo
13.33 kWh included; highest priority; overage at $7.50/kWh
 — No training
TrainingNo
Prompt retentionPartial
PrivacyTerms
76.2 Artificial Analysis Coding Index
Kimi K3
Kimi K3 (max)
 View
Ollama Cloud
HQ: San Francisco, CA, US
Models: glm-5.2, glm-5.1, deepseek-v4-pro, deepseek-v4-flash, kimi-k3, kimi-k2.7-code, kimi-k2.6, minimax-m3, minimax-m2.7, nemotron-3-ultra, nemotron-3-super, nemotron-3-nano, mistral-large-3, gpt-oss, gemma4, qwen3.5
Access: OpenAI-compatible
Free$0/mo
Cloud baseline
Pro$20/mo
50× Free; 3 con sessions
Max$100/mo
5× Pro; 10 con sessions; new sign-ups paused
Team$25/seat/mo
5-seat minimum ($125/mo); usage included per seat; shared billing and administration; US + EU regions; ZDR
 ZDR
TrainingNo
Prompt retentionZero
PrivacyCloud docs
76.2 Artificial Analysis Coding Index
Kimi K3
Kimi K3 (max)
 View
OpenCode
HQ: US (OpenCode)
Models: Grok 4.5, Kimi K3, Qwen3.8 Max, GLM-5.2, MiniMax M3, DeepSeek V4 Pro, GPT 5.6 Luna, Qwen3.7 Plus, Hy3, MiMo-V2.5, DeepSeek V4 Flash
Access: OpenCode-native, OpenAI-compatible tools
Go$10/mo
$5 first month; Go 120, Grok 4.5, Kimi K3 (2× usage), GLM-5.2, DeepSeek V4 Pro
 — — — Partial
TrainingPartial
Prompt retentionPartial
PrivacyGo docs
76.2 Artificial Analysis Coding Index
Kimi K3
Kimi K3 (max)
 View
Synthetic
HQ: US (Synthetic Lab Co.)
Models: moonshotai/Kimi-K3, nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-NVFP4, zai-org/GLM-5.2, zai-org/GLM-4.7-Flash, Qwen/Qwen3.6-27B, openai/gpt-oss-120b
Access: UI + API, agent frameworks, OpenAI-compatible
Subscription$30/mo
$1/d option; 500 msg/5h; 1 con/model/pack
 — — — ZDR
TrainingNo
Prompt retentionZero
Privacy
76.2 Artificial Analysis Coding Index
Kimi K3
Kimi K3 (max)
 View
Router9
HQ: Not stated
Models: OpenRouter catalog, anthropic/claude-opus-4.8, anthropic/claude-opus-4.8-fast, deepseek/deepseek-v4-pro, deepseek/deepseek-v4-flash, google/gemini-3.5-flash, minimax/minimax-m3, moonshotai/kimi-k2.7-code
Access: Claude Code, Cursor, OpenAI-compatible API, OpenClaw skills, MCP skills
Free$0/mo
50K credits/mo; 20 req/4h; free-tier models only; 7-day audit logs
Pro$29/mo
500K credits/mo; 500 req/4h; all models; 60% rollover up to 2×; 30-day audit logs
Max$59/mo
1.5M credits/mo; 2K req/4h; all models high-speed; 80% rollover up to 2×; unlimited audit logs
 — Standard
TrainingMay use
Prompt retentionLimited
PricingBilling docsAudit logsSecurity
74.3 Artificial Analysis Coding Index
Claude Opus 4.8
Claude Opus 4.8 (Adaptive Reasoning, Max Effort)
 View
Atlas Cloud Coding Plan
HQ: New York, NY, US
Models: Qwen3.8 Max, DeepSeek V4 Flash 0731, Doubao Seed 2.1 Pro, Doubao Seed 2.1 Turbo, GLM-5.2, Kimi K2.7 Code, MiniMax M3, DeepSeek V4 Pro, Kimi K2.6, Qwen3.6 Plus, GLM-5.1, MiniMax M2.7, MiniMax M2.5, GLM-5, Kimi K2.5, DeepSeek V3.2
Access: Claude Code, Cursor, OpenClaw, Codex CLI, Gemini CLI, OpenCode, Cline, Roo Code, Kilo Code, Trae, MCP, OpenAI-compatible API
Note: Included because monthly points are advertised as 35%-50% off Atlas Cloud pay-as-you-go pricing.
Starter$10/mo
16.5M points/week; 35%-50% off pay-as-you-go; coding-plan scoped API key
Lite$20/mo
33M points/week; 35%-50% off pay-as-you-go
Plus$50/mo
82.5M points/week; 35%-50% off pay-as-you-go
Max$100/mo
165M points/week; 35%-50% off pay-as-you-go
Ultra$200/mo
330M points/week; 35%-50% off pay-as-you-go
Enterprise$500/mo
825M points/week; SLA and custom options
 Standard
TrainingMay use
Prompt retentionVariable
Coding PlanDocsPrivacyDeletion
71.8 Artificial Analysis Coding Index
Qwen3.8 Max
Qwen3.8 Max
 View
AI Router
HQ: Switzerland
Models: Qwen3.6 (262K ctx), DeepSeek-V4-Flash, Qwen3-Embedding 4B (32K ctx)
Access: OpenAI-compatible API, OpenClaw, OpenCode, cURL, Python SDK
Flat~$48/mo
CHF 39/mo; fair-use unlimited; 3 parallel; 240 RPM; 10M TPM; personal/dev use
 — — — ZDR
TrainingNo
Prompt retentionZero
PrivacyTermsFair UseDPA
69.1 Artificial Analysis Coding Index
DeepSeek V4 Flash 0731
DeepSeek V4 Flash 0731 (Reasoning, Max Effort)
 View
Alibaba Cloud Model Studio — Token Plan
HQ: Hangzhou, China
Models: qwen3.7-max, qwen3.7-plus, qwen3.6-plus, qwen3.6-flash, deepseek-v4-pro, deepseek-v4-flash, kimi-k2.7-code, kimi-k2.6, glm-5.2, glm-5.1, glm-5, MiniMax-M2.5, qwen-image-2.0, wan2.7-image
Access: Mainstream coding tools, agent tools, OpenAI-compatible tools
Standard$30/mo
Per seat; 25K credits/seat/mo
Pro$100/mo
Per seat; 100K credits/seat/mo
Max$200/mo
Per seat; 250K credits/seat/mo
 — No training
TrainingNo
Prompt retentionVariable
Token Plan docsPrivacySecurity
69.1 Artificial Analysis Coding Index
DeepSeek V4 Flash
DeepSeek V4 Flash 0731 (Reasoning, Max Effort)
DeepSeek V4 Flash is the highest-scoring benchmarked model on the Alibaba Token Plan; DeepSeek V4 Pro and GLM-5.2 follow at 68.8.
 View
Arli AI
HQ: Indonesia
Models: DeepSeek-V4-Flash, Gemma-4-31B, GLM-4.7, GLM-4.6-Derestricted, Mistral-Medium-3.5, Qwen3.5-27B-Derestricted, 100+ community finetunes
Access: Roo Code, Kilo Code, OpenAI-compatible API, Image Generation
Plus$20/mo
Unlim tokens/req; 1 req at a time; up to 355B; 128K ctx; image gen max 3MP; Standard priority
Pro$30/mo
Unlim tokens/req; 2 parallel; up to 355B; 256K ctx; image gen max 6MP; High priority
Max$160/mo
Unlim tokens/req; 6 parallel; up to 355B; 512K ctx; image gen max 6MP; High priority
 — ZDR
TrainingNo
Prompt retentionZero
PrivacyTermsModels
69.1 Artificial Analysis Coding Index
DeepSeek V4 Flash 0731
DeepSeek V4 Flash 0731 (Reasoning, Max Effort)
 View
Entrim AI (OpenClaw Plans)
HQ: Šmarješke Toplice, Slovenia (EU)
Models: Qwen 3.6 35B-A3B, Qwen 3.6 27B, Gemma 4 31B, DeepSeek V4 Flash
Access: OpenAI-compatible API, OpenClaw
Qwen 3.6 35B-A3B$29/mo
Fast everyday coding; 864M in / 17.3M out tokens per day; 200 tok/s; 262K ctx; concurrency 1-8
Qwen 3.6 27B$49/mo
Accuracy-focused; 518M in / 10.4M out tokens/day; 120 tok/s; 262K ctx
Gemma 4 31B$49/mo
Code plus images and PDFs; 648M in / 13M out tokens/day; 150 tok/s; 254K ctx
DeepSeek V4 Flash$129/mo
1M context, single huge passes; 777M in / 15.6M out tokens/day; 180 tok/s; 524K ctx
 ZDR
TrainingNo
Prompt retentionZero
PrivacyTerms
69.1 Artificial Analysis Coding Index
DeepSeek V4 Flash
DeepSeek V4 Flash 0731 (Reasoning, Max Effort)
 View
Canopy Wave
HQ: Santa Clara, CA, US
Models: Coding Plan: GLM-5.2, Kimi K2.6, MiniMax M2.5, MiMo-V2.5, Unlimited Token Plan: Kimi K2.6 + MiniMax M2.5
Access: All major coding tools, OpenAI-compatible
Note: Kimi K3 listed as coming soon on the Canopy Wave model library.
Fast Bundle$15/mo
GLM-5.2, Kimi K2.6, MiniMax M2.5, MiMo-V2.5; 500 req/d; 10K req/mo; paid API limit 1 RPS / 6 RPM
Pro Bundle$30/mo
GLM-5.2, Kimi K2.6, MiniMax M2.5, MiMo-V2.5; 500 req/d; 10K req/mo; paid API limit 1 RPS / 6 RPM
Unlimited Token Plan 50M$15.99/mo
Kimi K2.6 + MiniMax M2.5; 50M high-speed tokens, then Basic Assurance: 1 RPS / 2 RPM / 10 RPH / 50 RPD
Unlimited Token Plan 200M$59.99/mo
Kimi K2.6 + MiniMax M2.5; 200M high-speed tokens, then Basic Assurance: 1 RPS / 8 RPM / 40 RPH / 200 RPD
Unlimited Token Plan 500M$159.99/mo
Kimi K2.6 + MiniMax M2.5; 500M high-speed tokens, then Basic Assurance: 1 RPS / 20 RPM / 100 RPH / 500 RPD
 ZDR
TrainingNo
Prompt retentionZero
PrivacyAgreementRate limits
68.8 Artificial Analysis Coding Index
GLM-5.2
GLM-5.2 (max)
 View
CheapestInference
HQ: San Francisco, CA, US
Models: Kimi K3, Kimi K2.7, Kimi K2.6, GLM 5.2, MiniMax M3, DeepSeek V4 Flash, MiMo v2.5
Access: OpenAI SDK, Anthropic SDK, Management API, x402 protocol
Note: Unlimited tokens within subscribed 8-hour windows. Each key handles one request at a time; add keys for parallel reques…
Core$14.99/mo
Per 8h block: DeepSeek V4 Flash + MiMo v2.5 (1M ctx); 1 concurrent req; $12.74/mo annual
Frontier$57/mo
Per 8h block: Kimi K2.7/K2.6, GLM 5.2, MiniMax M3; 1 concurrent req; $48.45/mo annual
Flagship$149/mo
Per 8h block: Kimi K3; sold out; $126.65/mo annual
 — ZDR
TrainingNo
Prompt retentionZero
PlansPrivacyTerms
68.8 Artificial Analysis Coding Index
GLM-5.2
GLM-5.2 (max)
 View
NanoGPT
HQ: Not stated
Models: GLM 5.2, GLM 5.1, Xiaomi MiMo V2.5 Pro, MiMo V2.5, Kimi K2.7 Code, MiniMax M3, DeepSeek V4 Pro, open-source text models, HiDream, Chroma, Z Image Turbo, Qwen Image
Access: Web chat, OpenAI-compatible API, Anonymous sessions
Note: Pro subscription is for personal, non-commercial use. Included flagship open-source models consume included input token…
Pro$12/mo
60M input tok/wk; 5% discount on other text models; included open-source text + image models; web + API; personal use
 — — — Partial
TrainingPartial
Prompt retentionPartial
Privacy
68.8 Artificial Analysis Coding Index
GLM-5.2
GLM-5.2 (max)
Closed models (GPT-5.6, Fable 5) are pay-as-you-go on NanoGPT.
 View
Novita AI Coding
HQ: San Francisco, CA, US
Models: GLM-5, Kimi K2.5, GLM-4.7, Minimax M2.1, DeepSeek V3.2, MiniMax M2.5, Qwen3.5-397B-A17B, MiniMax M2.7, DeepSeek V4 Pro
Access: Unified API, Novita Coding
Lite$19.90/mo
50M quota; 45 RPM; 50M TPM
Pro$49.90/mo
150M quota; 150 RPM; 50M TPM
Max$199.90/mo
750M quota; 450 RPM; 50M TPM
 — Partial
TrainingPartial
Prompt retentionVariable
PrivacyTerms
68.8 Artificial Analysis Coding Index
DeepSeek V4 Pro
DeepSeek V4 Pro 0813 (Reasoning, Max Effort)
 View
Regolo.ai
HQ: Frosinone, Italy
Models: apertus-70b, brick-v1-beta, gemma-4-31b, gpt-oss-120b, gpt-oss-20b, Llama-3.3-70B-Instruct, minimax-m2.5, mistral-small-4-119b, qwen3-coder-next, qwen3.5-122b, qwen3.5-9b, qwen3.6-27b, deepseek-v4-pro, mimo-v2-5-pro
Access: OpenAI-compatible API, serverless inference, Regolo dashboard
Core~$42/mo
€39/mo; up to 600M tokens/mo (20M tokens/d); Core Chat Models; email support
Boost~$96/mo
€89/mo; up to 1.5B tokens/mo (50M tokens/d); priority throughput and support
Pro~$183/mo
€169/mo; up to 3B tokens/mo (100M tokens/d); production-scale; flat monthly
Scale~$345/mo
€319/mo; up to 6B tokens/mo (200M tokens/d); enterprise, high-throughput 24/7
 ZDR
TrainingNo
Prompt retentionZero
PrivacyTerms
68.8 Artificial Analysis Coding Index
DeepSeek V4 Pro
DeepSeek V4 Pro 0813 (Reasoning, Max Effort)
DeepSeek V4 Pro (0813) is the highest-scoring benchmarked model on Regolo; GLM-5.2 ties at 68.8.
 View
Alibaba Cloud Model Studio — Coding Plan
HQ: Hangzhou, China
Models: qwen3.7-plus (vision), qwen3.6-plus (vision), kimi-k2.5 (vision), glm-5, MiniMax-M2.5, qwen3.5-plus (vision), qwen3-max-2026-01-23, qwen3-coder-next, qwen3-coder-plus, glm-4.7
Access: OpenCode, OpenAI-compatible tools
Pro$50/mo
6K req/5h; 45K/wk; 90K/mo
 — — — No training
TrainingNo
Prompt retentionLimited
PrivacySecurity
55.9 Artificial Analysis Coding Index
Qwen3.7 Plus
Qwen3.7 Plus
 View
Yolo-Auto
HQ: Not stated
Models: Qwen3.6-35B-A3B (35B total, 3B active; 128K ctx)
Access: OpenAI-compatible API
Starter$6/mo
Unlimited Qwen3.6-35B; 1 concurrent unit; 128K ctx; no daily cap
Builder$10/mo
Unlimited Qwen3.6-35B; 2 concurrent units; 128K ctx; no daily cap
Pro$15/mo
4 concurrent units; 256K ctx; priority on shared capacity
 — Partial
TrainingPartial
Prompt retentionPartial
Terms
41.9 Artificial Analysis Coding Index
Qwen3.6-35B-A3B
Qwen3.6 35B A3B (Reasoning)
 View
Claudin.io
HQ: Not stated
Models: Claudinio unified coding-agent model, 256K context
Access: Claudinio Code desktop app, Claude Code, Cursor, Cline, OpenAI-compatible API, Anthropic-compatible API
Note: Flat-rate subscription with a generous hourly anti-runaway ceiling that only stops rogue agents. Annual billing discoun…
Lite$9/mo
Unlimited; $1.00/h anti-runaway ceiling
Essential$19/mo
Unlimited; $2.00/h anti-runaway ceiling
Pro$39/mo
Unlimited; $4.00/h anti-runaway ceiling
Ultra$99/mo
Unlimited; $10.00/h anti-runaway ceiling
 Partial
TrainingPartial
Prompt retentionPartial
PrivacyTerms
— Artificial Analysis Coding Index
Claudinio unified coding-agent model
Not benchmarked
Claudinio routing model is not benchmarked as a standalone LLM by Artificial Analysis.
 View

IDEs & Coding Tools

17 providers
Brand Lite / Entry Pro / Standard Plus / Max Scale / Ultra Privacy Artificial Analysis Coding Index Action
Augment Code
HQ: Palo Alto, CA, US
Models: Claude Opus 5, Claude Fable 5, Claude Sonnet 5, Claude Opus 4.8, Claude Sonnet 4.6, Claude Haiku 4.5, GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.6 Luna, GPT-5.5, GPT-5.4, GPT-5.2, GPT-5.1, Gemini 3.1 Pro, Kimi K3, GLM 5.2 +2
Access: VS Code, JetBrains, Auggie CLI, Cosmos, code review
Note: Indie/Standard/Max per-developer plans replaced by a flat Business plan (up to 50 seats) + Enterprise; includes $100/mo…
Business$100/mo
Flat — no per-seat charge; $100/mo usage included; top-ups pay-as-you-go; up to 50 seats
EnterpriseCustom
Custom usage and top-ups; unlimited users; multi-region compute; CMEK/SIEM; SSO; audit trails
 — — No training
TrainingNo
Prompt retentionStored
SecurityPrivacyModels
78.3 Artificial Analysis Coding Index
GPT-5.6 Sol
GPT-5.6 Sol (xhigh)
GPT-5.6 Sol is the highest-scoring benchmarked model on Augment.
 View
Command Code
HQ: San Francisco, CA, US
Models: taste-1, Claude Fable 5, Claude Opus 5, Claude Opus 4.8, Claude Opus 4.7, Claude Sonnet 5, Claude Sonnet 4.6, Claude Haiku 4.5, GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.6 Luna, GPT-5.5, GPT-5.4, GPT-5.4 Mini, GPT-5.3 Codex, Gemini 3.6 Flash +28
Access: Command Code CLI, interactive CLI, headless mode, sandbox/background agents, Taste memory, skills, custom agents, commands, MCP, usage analytics
Note: Pricing docs list $10/$70/$80/$150/$300 monthly credits for Go / GOAT / Pro / Max / Max 20× plans. Extra top-up credits…
Go$1/mo
$10 credits/mo; ~15K requests; taste-1 model access; 29+ open-source models
GOAT$10/mo
$70 credits/mo; ~75K requests; taste-1 + GLM-5.2, GPT-5.6 Luna, Tencent Hy3, DeepSeek V4 Flash, and more
Pro$20/mo
$80 credits/mo; open-source + premium models; usage analytics
Max$100/mo
$150 credits/mo; ~219K requests; taste-1 + open-source + premium; deals up to ~$600 usage
Max 20×$200/mo
$300 credits/mo; highest rate limits; priority support
Teams$40/mo
Shared credit bucket; pooled usage; central admin; ~35K requests
 Partial
TrainingPartial
Prompt retentionPartial
PricingPricing docsModelsSecurityPrivacyTerms
78.3 Artificial Analysis Coding Index
GPT-5.6 Sol
GPT-5.6 Sol (xhigh)
GPT-5.6 Sol is the highest-scoring benchmarked model; premium Claude models are on paid tiers.
 View
Cursor
HQ: San Francisco, CA, US
Models: Claude Fable 5, Claude Opus 5, Claude Sonnet 5, GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.6 Luna, Gemini 3.6 Flash, Gemini 3.1 Pro, Grok 4.5
Access: Cursor editor, coding agent, Composer, cloud agents, Bugbot
Hobby$0/mo
Limited Agent requests; Access to Composer
Individual$20/mo
Extended Agent limits; Grok & Composer generous limits; frontier models; MCPs/skills/hooks; cloud agents; Bugbot
Pro+$60/mo
3× Pro limits on Agent; Grok & Composer; frontier models; cloud agents
Ultra$200/mo
20× Pro limits; priority access to new features
Teams$40/user/mo
Centralized team billing; team marketplace; Bugbot code review; team privacy mode; SAML/OIDC SSO
 Partial
TrainingPartial
Prompt retentionPartial
Privacy overviewPrivacy mode
78.3 Artificial Analysis Coding Index
GPT-5.6 Sol
GPT-5.6 Sol (xhigh)
GPT-5.6 Sol is the highest-scoring frontier model on Cursor; Claude Opus 5 and Grok 4.5 also available.
 View
GitHub Copilot
HQ: San Francisco, CA, US
Models: GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.6 Luna, GPT-5.5, GPT-5.4, GPT-5.3 Codex, Claude Opus 5, Claude Fable 5, Claude Opus 4.8, Claude Opus 4.7, Claude Sonnet 5, Claude Sonnet 4.6, Gemini 3.6 Flash, Gemini 3.1 Pro, Grok 4.5, Kimi K3
Access: VS Code, JetBrains, Visual Studio, Neovim, Xcode, Eclipse, CLI
Free$0/mo
50 agent/chat req; 2K completions/mo; limited auto-mode agent
Pro$10/mo
$15 AI Credits ($10 base + $5 flex); unlimited completions
Pro+$39/mo
$70 AI Credits ($39 base + $31 flex); unlimited completions
Max$100/mo
$200 AI Credits ($100 base + $100 flex); unlimited completions; highest usage
 Standard
TrainingMay use
Prompt retentionVariable
Copilot dataPrivacyModelsBilling change
78.3 Artificial Analysis Coding Index
GPT-5.6 Sol
GPT-5.6 Sol (xhigh)
GPT-5.6 Sol is the highest-scoring benchmarked model available on Copilot (Claude Fable 5 remains available and may need admin enablement).
 View
Kilo / KiloClaw
HQ: San Francisco, CA, US / Amsterdam, NL
Models: Claude (incl. Opus 5 / Sonnet 5), GPT-5.6, Grok 4.5, Grok Build 0.1, Gemini 3.6 Flash, Kimi K3, GLM-5.2, MiniMax M3, DeepSeek V4 Pro, MiMo-V2.5-Pro, Qwen, Laguna S 2.1, Ling, Devstral 2
Access: VS Code, JetBrains, CLI, Kilo Pass, KiloClaw/OpenClaw, OpenRouter, Vercel, AWS Bedrock, Azure, local models, Cloud agents (per-hour)
Note: Restructured into Platform (Free / Teams $15/user / Enterprise) separate from the optional Kilo Pass monthly-credit sub…
Free (Platform)$0/mo
BYOK or pay provider rates; 500+ models across 60+ providers; Auto Model routing
Teams (Platform)$15/user/mo
Team management; usage analytics; shared agent modes; 14-day trial
Kilo Pass Starter$19/mo
Up to $26.60 credits/mo; up to 50% bonus credits
Kilo Pass Pro$49/mo
Up to $68.60 credits/mo; bonus credits
Kilo Pass Expert$199/mo
Up to $278.60 credits/mo; bonus credits
 Partial
TrainingPartial
Prompt retentionPartial
PricingPrivacyGateway data
78.3 Artificial Analysis Coding Index
GPT-5.6 Sol
GPT-5.6 Sol (xhigh)
GPT-5.6 Sol is the highest-scoring benchmarked model in Kilo's catalog.
 View
Kiro
HQ: Seattle, WA, US
Models: Auto (frontier mix), Claude Opus 5, Claude Sonnet 5, Claude Opus 4.8, Claude Opus 4.7, Claude Opus 4.6, Claude Opus 4.5, Claude Sonnet 4.6, Claude Sonnet 4.5, Claude Haiku 4.5, GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.6 Luna, DeepSeek 3.2, MiniMax M2.5, MiniMax M2.1 +2
Access: Kiro IDE
Free$0/mo
50 credits; 500 signup bonus credits for 30 days
Pro$20/mo
1K credits; pay-per-use overage
Pro+$40/mo
2K credits; pay-per-use overage
Pro Max$100/mo
5K credits; pay-per-use overage
Power$200/mo
10K credits; pay-per-use overage
 Standard
TrainingMay use
Prompt retentionVariable
Data protectionPrivacyModels
78.3 Artificial Analysis Coding Index
GPT-5.6 Sol
GPT-5.6 Sol (xhigh)
GPT-5.6 Sol is the highest-scoring benchmarked model on Kiro; Auto routes across the frontier set.
 View
Warp
HQ: New York, NY, US
Models: Claude Fable 5, Claude Opus 5, Claude Opus 4.8, Claude Sonnet 5, Claude Sonnet 4.6, Claude Haiku 4.5, GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.6 Luna, GPT-5.5, GPT-5.4, GPT-5.3 Codex, Gemini 3.5 Flash, Gemini 3.1 Pro, Grok 4.5, Kimi K3 +3
Access: Warp terminal, Warp Agent, Oz Agent Platform, cloud agents
Note: Pivoted to an agent platform (Oz Agent Platform, Warp Agent, cloud agents). Free tier now PAYG-reload-credit based rath…
Free$0/mo
Modern terminal; Warp Agent CLI access; PAYG reload credits; bring your own AI inference; limited cloud agents
Build$20/mo
Pay-as-you-go from $20; 1.5K credits ($20 included agent usage); full Warp Agent access; reload credits; extended cloud…
Business$50/user/mo
Per user (up to 25 seats); 1.5K credits/seat; team usage metrics; admin controls; shared Slack channel
Max$200/mo
18K credits (12× Build); reload credits; volume discounts; auto-reload; team spend cap; private email support
 ZDR
TrainingNo
Prompt retentionZero
ZDR docsPrivacy
78.3 Artificial Analysis Coding Index
GPT-5.6 Sol
GPT-5.6 Sol (xhigh)
GPT-5.6 Sol is the highest-scoring benchmarked model in Warp's curated set.
 View
Windsurf / Codeium
HQ: Mountain View, CA, US
Models: SWE-1.7, SWE-1.6, SWE-1.5, Claude Fable 5, Claude Opus 5, Claude Opus 4.8, Claude Sonnet 5, Claude Sonnet 4.6, Claude Haiku 4.5, GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.6 Luna, GPT-5.5, GPT-5.4, GPT-5.3, Gemini 3.6 Flash +7
Access: Windsurf editor, Cascade, Tab, Devin Cloud agents, Devin Desktop
Note: Cognition merged Windsurf/Codeium with Devin; cloud agents now run via Devin Cloud/Devin Desktop. SWE-1.7 is the in-hou…
Free$0/mo
Light quota; limited model availability; unlimited inline edits and Tab
Pro$20/mo
Increased quotas; OpenAI/Claude/Gemini frontier; full models; free SWE 1.7 + open source; Devin Cloud agents; extra usa…
Max$200/mo
Significantly higher quotas
Teams$80/mo + $40/seat
Team plan + per full dev seat; unlimited members; central billing; admin dashboard with analytics; priority support
 Partial
TrainingPartial
Prompt retentionPartial
PrivacySecurityModels
78.3 Artificial Analysis Coding Index
GPT-5.6 Sol
GPT-5.6 Sol (xhigh)
Highest-scoring benchmarked frontier model on Windsurf; in-house SWE-1.7 flagship is not benchmarked by Artificial Analysis.
 View
Zed
HQ: Denver, CO, US
Models: Claude Fable 5, Claude Opus 5, Claude Opus 4.8, Claude Opus 4.7, Claude Opus 4.6, Claude Opus 4.5, Claude Opus 4.1, Claude Sonnet 5, Claude Sonnet 4.6, Claude Sonnet 4.5, Claude Sonnet 4, Claude Haiku 4.5, GPT-5.6, GPT-5.5 pro, GPT-5.5, GPT-5.4 pro +15
Access: Zed editor, Agent Panel, Inline Assistant, Edit Prediction, External agents
Personal$0/mo
2K accepted edit predictions; external agents
Pro$10/mo
Unlimited edit predictions; $5 tokens included; usage-based billing beyond included credits
Business$30/mo
Per seat; org AI model policies; data governance controls; unified spend visibility
 — ZDR
TrainingNo
Prompt retentionZero
PrivacyTerms
78.3 Artificial Analysis Coding Index
GPT-5.6 Sol
GPT-5.6 Sol (xhigh)
GPT-5.6 Sol is the highest-scoring benchmarked model on Zed.
 View
ClinePass
HQ: San Francisco, CA, US
Models: GLM 5.2, Kimi K3, Kimi K2.7 Code, Kimi K2.6, DeepSeek V4 Pro, DeepSeek V4 Flash, MiniMax M3, MiMo V2.5 Pro, MiMo V2.5, Qwen3.7 Max, Qwen3.7 Plus
Access: Cline IDE, Cline CLI, Cline SDK, Other agents via API key
ClinePass$9.99/mo
$4.99 first month promo; 2-5× API rate usage; 5h rolling + weekly + monthly limits
 — — — Partial
TrainingPartial
Prompt retentionVariable
ClinePassDocsPrivacy
76.2 Artificial Analysis Coding Index
Kimi K3
Kimi K3 (max)
 View
Tabnine
HQ: Tel Aviv, Israel
Models: Tabnine proprietary code models, Claude 4.8 Opus, Claude 4.7 Opus, Claude 4.6 Sonnet, Claude 4.6 Opus, Claude 4.5 Sonnet, Claude 4.5 Opus, Claude 4.5 Haiku, Claude 4 Sonnet, GPT-5.5, GPT-5.4, GPT-5.3 Codex, GPT-5.2 Codex, GPT-5.2, GPT-5, Gemini 3.5 Pro +9
Access: VS Code, JetBrains, CLI, MCP
Note: Both tiers are sold as annual subscriptions; per-month equivalent shown. Deployment options include SaaS, VPC, on-prem,…
Code Assistant Platform$39/mo
Per user; annual billing; IDE assistant; Tabnine LLM at cost + 5%
Agentic Platform$59/mo
Per user; annual billing; autonomous agents; CLI; MCP
 — — ZDR
TrainingNo
Prompt retentionZero
Code privacyPrivacy
74.9 Artificial Analysis Coding Index
GPT-5.5
GPT-5.5 (xhigh)
 View
Zencoder
HQ: Campbell, CA, US
Models: GPT-5.5, GPT-5.4, GPT-5.4 Mini, GPT-5.3 Codex, Claude Opus 4.7, Claude Opus 4.6, Claude Sonnet 4.6, Claude Haiku 4.5, Gemini Pro 3.1, Gemini Flash 3.0, Grok Code Fast 1, Auto, Auto+
Access: Zenflow Desktop, IDE Agents, VS Code, JetBrains, Android Studio, Autonomous Agents, multi-model orchestration, MCP, multi-repository indexing
Note: Credits are consumed by model/work; docs give Claude Sonnet 4.6 typical request at 100-500 credits, average 250-300. Ef…
Pro$45/mo
30K credits/mo; $40/mo annual; IDE Agents; Zenflow Desktop
Pro Plus$95/mo
80K credits/mo; $85/mo annual; shared team pool; multi-repo indexing; SSO/audit logs
Pro Max$195/mo
180K credits/mo; $175/mo annual; priority support
EnterpriseCustom
Prepaid usage plans; unlimited multi-repo indexing; private deployment; professional services
 Partial
TrainingPartial
Prompt retentionPartial
PricingPricing docsModelsSecurityPrivacyTermsTrust
74.9 Artificial Analysis Coding Index
GPT-5.5
GPT-5.5 (xhigh)
 View
Codebuff
HQ: Not stated
Models: Claude Opus 4.7, GPT-5.6 Luna, GPT-5.4, GPT-5.1, Gemini 3.1 Flash Lite, Gemini 2.0 Flash
Access: Codebuff CLI, Default mode, Max mode, Plan mode, Lite mode, parallel agents, code review, terminal commands, file discovery, knowledge files, MCP, custom agents +1
Note: Pricing page lists PAYG at $0.01/credit and Strong subscription tiers at $100/$200/$500 for 1×/2.5×/7× usage. Credit to…
Signup credits$0/mo
500 free credits on signup; credits consumed by task complexity
Strong 1×$100/mo
16,800 credits/mo; 168 credits/$; $168 PAYG-equivalent at $0.01/credit
Strong 2.5×$200/mo
42,000 credits/mo; 210 credits/$; $420 PAYG-equivalent at $0.01/credit
Strong 7×$500/mo
117,500 credits/mo; 235 credits/$; $1,175 PAYG-equivalent at $0.01/credit
 No training
TrainingNo
Prompt retentionVariable
PricingModelsModesLarge projectsPrivacyFAQ
73.6 Artificial Analysis Coding Index
Claude Opus 4.7
Claude Opus 4.7 (Adaptive Reasoning, Max Effort)
Default/Max use Claude Opus 4.7; Lite uses GPT-5.6 Luna.
 View
Trae
HQ: Singapore / ByteDance
Models: Gemini 3.1 Pro, Gemini 3 Flash, GPT-5 family, GPT-5.4, GPT-5.2, MiniMax M3, MiniMax M2.7, Kimi, DeepSeek R1, DeepSeek V3
Access: Coding agent, cloud tasks
Free$0/mo
5K autocomplete/mo; limited AI req
Lite$3/mo
$5 usage value; 2 con cloud tasks
Pro$10/mo
$20 usage value; 10 con cloud tasks
Pro+$30/mo
$90 usage value; 15 con cloud tasks
Ultra$100/mo
$400 usage value; 20 con cloud tasks
 Partial
TrainingPartial
Prompt retentionVariable
PrivacyPrivacy Mode
71.1 Artificial Analysis Coding Index
GPT-5.4
GPT-5.4 (xhigh)
 View
Google Antigravity
HQ: Mountain View, CA, US
Models: Gemini 3.5 Flash, Gemini 3.1 Pro, Gemini 3 Flash, Claude Sonnet 4.6, Claude Opus 4.6, gpt-oss-120b
Access: Agentic coding, browser, validation, Google Cloud org plan (Antigravity 2.0 + CLI)
Note: Google AI Plus $7.99 tier is no longer listed; paid access now via Google AI Pro / Ultra (also used for Antigravity 2.0…
Individual$0/mo
$0/mo; no subscription; unlimited Tab completions and command requests; basic weekly rate limits
Google AI Pro$19.99/mo
Via Google AI subscription; more generous rate limits; flexible AI credit pool
Google AI Ultra$99.99/mo
Via Google AI Ultra; higher access to latest Gemini models; more generous limits
 — Partial
TrainingPartial
Prompt retentionVariable
TermsGoogle privacyPricing
70.1 Artificial Analysis Coding Index
Gemini 3.5 Flash
Gemini 3.5 Flash (high)
Gemini 3.5 Flash is the highest-scoring listed Gemini model (above Gemini 3.1 Pro).
 View
Verboo Code
HQ: Fortaleza, CE, Brazil
Models: qwen3.6-27b, glm-4.7-flash, deepseek-v4-flash, mimo-v2.5, minimax-m3, deepseek-v4-pro, mimo-v2.5-pro, kimi-k2.7, glm-5.2
Access: Verboo Code CLI, OpenAI-compatible API, VS Code, Cursor, Terminal, Pi Agent, OpenClaw, Hermes Agent
Note: BRL-only pricing (Pix), converted at the site's stated ~6 BRL/USD. All plans unlimited tokens with 30 req/min and 2 con…
Junior~$12/mo
R$75/mo; unlimited tokens; 30 req/min; 2 con; qwen3.6-27b + glm-4.7-flash
Pro~$20/mo
R$119/mo; unlimited tokens; 30 req/min; 2 con; adds deepseek-v4-flash, mimo-v2.5; 1-day free trial
Max~$53/mo
R$319/mo; unlimited tokens; 30 req/min; 2 con; adds minimax-m3, deepseek-v4-pro, mimo-v2.5-pro
Ultra~$150/mo
R$899/mo; unlimited tokens; 30 req/min; 2 con; adds kimi-k2.7, glm-5.2 (all 9 models)
 No training
TrainingNo
Prompt retentionPartial
PlansGitHubPrivacyTerms
69.1 Artificial Analysis Coding Index
DeepSeek V4 Flash 0731
DeepSeek V4 Flash 0731 (Reasoning, Max Effort)
 View
Amazon Q Developer
HQ: Seattle, WA, US
Models: Amazon Q Developer model, Claude (latest models)
Access: VS Code, JetBrains, Visual Studio, Eclipse, CLI, AWS tooling
Free$0/mo
50 agentic req/mo; 1K LOC transformations
Pro$19/mo
Per user; higher limits; 4K LOC transformations/user/mo
 — — Partial
TrainingPartial
Prompt retentionVariable
Data protectionData storage
— Artificial Analysis Coding Index
Amazon Q Developer model
Not benchmarked
Not benchmarked by Artificial Analysis.
 View

Autonomous Agents

4 providers
Brand Lite / Entry Pro / Standard Plus / Max Scale / Ultra Privacy Artificial Analysis Coding Index Action
Factory.ai
HQ: San Francisco, CA, US
Models: Claude Fable 5, Claude Opus 5, Claude Opus 4.8, Claude Opus 4.7, Claude Opus 4.6, Claude Opus 4.5, Claude Sonnet 5, Claude Sonnet 4.6, Claude Sonnet 4.5, Claude Haiku 4.5, GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.6 Luna, GPT-5.5, GPT-5.4, GPT-5.3 +17
Access: Droid, Desktop, CLI, SDK, cloud agents, local agents
Pro$20/mo
Desktop / CLI / SDK; cloud & local background agents
Plus$100/mo
~5× Pro usage; Droid Computers cloud computers
Max$200/mo
~10× Pro usage; early access to new features
 — No training
TrainingNo
Prompt retentionVariable
PrivacyData flowsModels
78.3 Artificial Analysis Coding Index
GPT-5.6 Sol
GPT-5.6 Sol (xhigh)
GPT-5.6 Sol is the highest-scoring benchmarked model on Factory.
 View
Jules (Google)
HQ: Mountain View, CA, US
Models: Gemini 3.5 Flash, Gemini 3 Pro, Gemini 2.5 Pro
Access: Async coding, GitHub
Free (via Google AI)$0/mo
15 tasks/d; 3 con; Gemini 2.5 Pro
Google AI Pro$19.99/mo
100 tasks/d; 15 con; starts Gemini 3 Pro
Google AI Ultra$99.99/mo
300 tasks/d; 60 con; starts Gemini 3 Pro
 — Partial
TrainingPartial
Prompt retentionVariable
Google privacyJules limits
70.1 Artificial Analysis Coding Index
Gemini 3.5 Flash
Gemini 3.5 Flash (high)
Jules agent runs Gemini 3.5 Flash as its flagship model.
 View
Cosine (Genie)
HQ: London, UK
Models: Lumen Scout (codebase intelligence), Lumen Outpost (production coding), Lumen Sovereign (coming soon)
Access: Autonomous coding, Claude/OpenAI/Copilot CLI passthrough
Note: Per-seat Hobby/Professional plans replaced by credit-based Starter/Team/Enterprise with add-on credit pricing. Credits…
Starter$19/mo
4M credits/mo; $6.50 per 1M add-on credits; solo devs / side projects
Team$199/mo
47M credits/mo; $5.00 per 1M add-on credits; growing teams
Enterprise$999/mo
240M credits/mo; $4.50 per 1M add-on credits; highly-regulated industries
 — No training
TrainingNo
Prompt retentionPartial
PrivacySecurity FAQ
— Artificial Analysis Coding Index
Lumen (Scout / Outpost / Sovereign)
Not benchmarked
Cosine's Lumen model family is not benchmarked as a standalone LLM by Artificial Analysis.
 View
Devin
HQ: San Francisco, CA, US
Models: SWE 1.7 (Cognition), Devin autonomous software engineer
Access: Autonomous coding, Cloud tasks, Devin Review, DeepWiki, Windsurf IDE, Slack, Linear, MCP
Note: Cognition merged Devin with Windsurf/Codeium; flagship software-engineering model is now SWE 1.7. Pricing/plan structur…
Free$0/mo
Light quota; limited models; unlimited inline edits and Tab; Devin Review; DeepWiki
Pro$20/mo
Increased quotas; OpenAI/Claude/Gemini frontier; full models; free SWE 1.7 + open source; Devin Cloud; Slack/Linear/MCP…
Max$200/mo
Significantly higher quotas; up to 10 con
Teams$80/mo + $40/seat
Team plan + per full dev seat; unlimited members; central billing; admin dashboard; priority support
 Standard
TrainingMay use
Prompt retentionVariable
PrivacySecurity
— Artificial Analysis Coding Index
Devin autonomous software engineer
Not benchmarked
Agent product, not benchmarked as a standalone LLM by Artificial Analysis.
 View

App Builders

8 providers
Brand Lite / Entry Pro / Standard Plus / Max Scale / Ultra Privacy Artificial Analysis Coding Index Action
VULK
HQ: Vila do Conde, Porto, Portugal
Models: Claude Fable 5, Claude Opus 4.8, Claude Opus 4.7, Claude Sonnet 4.6, GPT-5.5, GPT-5.3 Codex, Gemini 3.5 Flash, Gemini 3.1 Pro, Gemini 3.1 Flash, Grok 4.20, Grok 4.1, GLM-5.2, MiniMax M2.7, DeepSeek, Qwen3 Coder, Llama 3.3 +1
Access: Web app builder, Flutter mobile
Note: Team and Business tiers removed from the pricing page; only Builder / Pro / Max are offered. Pro adds AI video + 3D Stu…
Builder~$22/mo
EUR 19.99/mo; 1,000 credits/mo; all models; production deploy; custom domains; ZIP export; backend + Android APK
Pro~$44/mo
EUR 39.99/mo; 2,500 credits/mo; image/video gen; 3D Studio; BYOM
Max~$220/mo
EUR 199/mo; 10,000 credits/mo; power users + agencies; priority routing; full API/SDK; app store publishing
 — Partial
TrainingPartial
Prompt retentionStored
PrivacyTerms
76.5 Artificial Analysis Coding Index
Claude Fable 5
Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)
 View
cto.new
HQ: London, UK
Models: Claude Opus 4.8, GPT-5.5, Claude Sonnet 4.6, Gemini 3.5 Flash, Gemini 3.1 Pro, GLM 5.1, GPT 5.3 Codex, Grok 4.20, Kimi K2.6, MiniMax M2.7
Access: Browser app builder, CTOClaw, CLI, MCP
Note: Premium starts at $20/mo. Published usage limits are rolling 24-hour and 7-day limits, but exact token allotments and c…
Free$0/mo
Daily-reset usage; best-value models; web/mobile apps; CTOClaw agents; code-agent tasks; API/MCP integrations
Premium$20/mo
More tokens; wider high-performance model access; rolling 24h and 7d limits
 — — Standard
TrainingMay use
Prompt retentionStored
PricingPremium docsFree docsModelsAgentDataPrivacyTrust
74.9 Artificial Analysis Coding Index
GPT-5.5
GPT-5.5 (xhigh)
Auto model routing varies.
 View
Bolt.new
HQ: San Francisco, CA, US
Models: Standard, Sonnet 4.6, Opus 4.6, Opus 4.7
Access: App building
Free$0/mo
1M tok/mo; 300K daily cap
Pro$25/mo
10M tok/mo; no daily cap; 1-month rollover
Teams$30/mo
Per member; centralized billing/admin; org sharing; private NPM
 — Standard
TrainingMay use
Prompt retentionStored
73.6 Artificial Analysis Coding Index
Claude Opus 4.7
Claude Opus 4.7 (Adaptive Reasoning, Max Effort)
 View
a0.dev
HQ: California, US
Models: Opus 4.7, GPT 5.4, Advanced coding models (names not published)
Access: Mobile app builder
Pro$20/mo
100 credits/mo; advanced coding models; unlimited app builds; custom domain; no branding; code export
Max$200/mo
1,250 credits/mo; top AI models; 250 bonus credits/mo; unlimited support
 — — Standard
TrainingMay use
Prompt retentionStored
Privacy
71.1 Artificial Analysis Coding Index
GPT 5.4
GPT-5.4 (xhigh)
 View
Base44
HQ: Not stated
Models: Base44 app builder, AI model select (model names not published), Superagents, Generated frontend/backend
Access: Browser app builder, GitHub integration
Note: Model names and context window are not published. Integration credits are consumed by app-side integrations; each integ…
Free$0/mo
25 message credits/mo; core features; 100-500 integration credits/mo listed
Starter$16/mo
Billed annually; 100 message credits/mo; 2K integration credits; in-app code edits
Builder$40/mo
Billed annually; 250 message credits/mo; 10K integration credits; backend functions; model select
Pro$80/mo
Billed annually; 500 message credits/mo; 20K integration credits; GitHub integration; beta access
Elite$160/mo
Billed annually; 1,200 message credits/mo; 50K integration credits; premium support
 Standard
TrainingMay use
Prompt retentionVariable
PricingPrivacyTermsDPASuperagents
— Artificial Analysis Coding Index
Base44 app builder (models not named)
Not benchmarked
Base44 publishes AI model selection but does not name model IDs or context windows; app-builder product is not benchmarked as a standalone LLM.
 View
Lovable
HQ: Stockholm, Sweden
Models: Lovable app builder, Code mode (paid)
Access: App building
Free$0/mo
5 daily credits; up to 30/mo
Pro$25/mo
100 monthly credits + 5 daily up to 150/mo
Business$50/mo
100 monthly credits; SSO; team workspace; role-based access; security center
 — Partial
TrainingPartial
Prompt retentionStored
— Artificial Analysis Coding Index
Lovable app builder
Not benchmarked
App-builder product, not benchmarked as a standalone LLM by Artificial Analysis.
 View
Replit
HQ: San Francisco, CA, US
Models: Replit Agent
Access: Cloud IDE, App building
Starter$0/mo
Daily Agent credits; 1 published app
Core$25/mo
$25 monthly credits; 5 collaborators
Pro$100/mo
$100 monthly credits; 15 collaborators; private deployments
 — Partial
TrainingPartial
Prompt retentionVariable
PrivacyAI data
— Artificial Analysis Coding Index
Replit Agent
Not benchmarked
Agent product, not benchmarked as a standalone LLM by Artificial Analysis.
 View
v0 / Vercel
HQ: San Francisco, CA, US
Models: v0 Mini, v0 Pro, v0 Max, v0 Max Fast
Access: App building
Free$0/mo
$5 monthly credits; 7 messages/day; 128K ctx; 32K output
Plus$30/mo
Per user; access to all models; $30 monthly credits/user; $2 daily login credits; shared credit pool
Business$100/mo
Per user; access to all models; $30 monthly credits/user; $2 daily login credits; training opt-out by default; SAML SSO…
 — Partial
TrainingPartial
Prompt retentionStored
AI termsPrivacy
— Artificial Analysis Coding Index
v0 Max
Not benchmarked
v0 model family is not benchmarked by Artificial Analysis.
 View

Deprecated / Closed

14 providers
Brand Lite / Entry Pro / Standard Plus / Max Scale / Ultra Privacy Artificial Analysis Coding Index Action
BLACKBOX AI
HQ: San Francisco, CA, US
Models: GPT-5.5, Claude Opus 4.7, Claude Sonnet 4.6, Gemini 3.1 Pro, GPT-5.3 Codex, Kimi K2.6, MiniMax M2.7, Qwen3 Coder, GLM 5, Blackbox Pro, 400+ model catalog
Access: BLACKBOX IDE, VS Code extension, CLI, API, Cloud agents, Builder, Mobile, multi-agent execution, MCP
Note: Deprecated as a subscription coding plan. Subscription tiers (Pro/Pro Plus/Pro Max) have been removed; Blackbox now off…
Free$0/mo
Unlimited agent requests with MiniMax M2.5; basic IDE/API access
Pro$10/mo
$20 model credits; all chat models; voice agent; unlimited MiniMax M2.5 agent requests
Pro Plus$20/mo
$40 model credits; multi-agent execution; app builder; coding agent across IDEs, web, and terminal
Pro Max$40/mo
$80 model credits; unlimited agent requests; team controls, SAML SSO, analytics, priority support
 Partial
TrainingPartial
Prompt retentionPartial
PricingAPISecurityZDR docsPrivacy
74.9 Artificial Analysis Coding Index
GPT-5.5
GPT-5.5 (xhigh)
Deprecated; subscription plans removed, now pay-as-you-go + enterprise only.
 View
SwiftRouter
HQ: Not stated
Models: gpt-5.5 (1.05M ctx), gpt-5.4 (1.05M ctx), gpt-5.3-codex (400K ctx), claude-opus-4.7 (1M ctx), claude-sonnet-4.6 (1M ctx), qwen3-coder-480b (262K ctx), devstral-2 (262K ctx), deepseek-v4-flash (1M ctx), gemini-3.1-pro-preview (1M ctx), 41 starter models / public catalog
Access: OpenAI-compatible API, Claude Code, Codex, OpenCode, Kilocode, RooCode, Cline, OpenAI SDK clients, Playground
Note: Paid plans are fixed-term one-month purchases and do not auto-renew. Included plan credits post once after payment conf…
Starter$0/mo
$10 free Universal Credit; 41 starter models; 5 RPM; 1M tokens/d; 7M tokens/wk
Lite$10/mo
$30 Universal + $50 Swift credit; 20 RPM; all models; 5M tokens/d; 35M tokens/wk
Pro$20/mo
$50 Universal + $80 Swift credit; 60 RPM; all models; unlimited token quota
Max$50/mo
$100 Universal + $150 Swift credit; 150 RPM; all models; unlimited token quota
 Partial
TrainingPartial
Prompt retentionStored
PricingDocsAgentsPrivacyTerms
74.9 Artificial Analysis Coding Index
GPT-5.5
GPT-5.5 (xhigh)
 View
CrofAI
HQ: US (Nahcrof LLC)
Models: DeepSeek V4 Pro, DeepSeek V4 Pro (Precision), DeepSeek V4 Flash, DeepSeek V3.2, GLM 5.1, GLM 5.1 (Precision), Kimi K2.6, Kimi K2.6 (Precision), Kimi K2.5, Kimi K2.5 Lightning, GLM 5, GLM 4.7, GLM 4.7 Flash, Gemma 4 31B, MiniMax M2.5, Qwen3.6 27B +1
Access: OpenAI-compatible tools
Note: Subscriptions removed May 31 2026. Service moved to pay-as-you-go only.
Hobby$5/mo
500 req/d
Pro$10/mo
1K req/d
Intermediate$20/mo
2.5K req/d
Scale$50/mo
6.25K req/d
Max$100/mo
12.5K req/d
Flexible$20/mo+
125 req/d per $1/mo; pick your price ($20+)
 ZDR
TrainingNo
Prompt retentionZero
Privacy
68.8 Artificial Analysis Coding Index
DeepSeek V4 Pro
DeepSeek V4 Pro 0813 (Reasoning, Max Effort)
 View
Featherless.ai
HQ: San Francisco, CA, US
Models: deepseek-ai/DeepSeek-V4-Pro, deepseek-ai/DeepSeek-V3.2, zai-org/GLM-5.2, zai-org/GLM-5.1, zai-org/GLM-5, zai-org/GLM-4.7, MiniMaxAI/MiniMax-M2.5, moonshotai/Kimi-K2.5, Qwen/Qwen3-235B-A22B
Access: OpenClaw (Claw tiers), Roo Code, Kilo Code
Agent Standard$100/mo
Unlim; 8 con; 256K ctx; models up to 229B; sandbox runtime
Agent Pro$200/mo
Unlim; 8 con; 256K ctx; any catalog model; larger sandbox
 — — ZDR
TrainingNo
Prompt retentionZero
PrivacyLogging
68.8 Artificial Analysis Coding Index
GLM-5.2
GLM-5.2 (max)
Deprecated; no longer qualifies as a subscription plan.
 View
Umans Code
HQ: Montrouge, France
Models: GLM-5.2, Kimi K2.7-Code, Qwen3.6-35B-A3B-FP8
Access: Claude Code, Cursor, Crush, OpenCode, Kilo Code, Cline, TRAE, Grok CLI, Pi, OpenAI-compatible, Anthropic-compatible
Note: Beta status. Drop-in Claude Code backend replacement using open-source models. Subscription plans deprecated; now pay-p…
Pro$20/mo
Up to 200 messages per 5h window; rolling reset; up to 5 concurrent requests; no weekly limits
 — — — ZDR
TrainingNo
Prompt retentionZero
PrivacyTerms
68.8 Artificial Analysis Coding Index
GLM-5.2
GLM-5.2 (max)
 View
Umans Code Max
HQ: Montrouge, France
Models: GLM-5.2, Kimi K2.7-Code, Qwen3.6-35B-A3B-FP8
Access: Claude Code, Cursor, Crush, OpenCode, Kilo Code, Cline, TRAE, Grok CLI, Pi, OpenAI-compatible, Anthropic-compatible
Note: $50 Max plan is deprecated / waitlist and no longer generally available.
Max$50/mo
Deprecated / waitlist; unlimited prompts; up to 4 concurrent sessions; burst capacity
 — — — No training
TrainingNo
Prompt retentionVariable
PrivacyTerms
68.8 Artificial Analysis Coding Index
GLM-5.2
GLM-5.2 (max)
Deprecated / waitlist Umans Code Max listing.
 View
Wafer Pass
HQ: San Francisco, CA, US
Models: Qwen3.5-397B-Turbo, GLM5.1-Turbo, DeepSeekV4-Pro-Turbo
Access: OpenAI-compatible tools
Note: Wafer Pass has shut down; plans are no longer available.
Lite$12/mo
100 req/5h
Starter$40/mo
1K req/5h
Privacy$100/mo
2K req/5h; explicit ZDR
 — Partial
TrainingPartial
Prompt retentionPartial
PrivacyTerms
68.8 Artificial Analysis Coding Index
DeepSeek V4 Pro
DeepSeek V4 Pro 0813 (Reasoning, Max Effort)
 View
routing.sh
HQ: India (Routing Run Private Limited)
Models: Qwen3.6 27B 202K, DeepSeek V4 Pro, DeepSeek V4 Flash, DeepSeek V3.2, Kimi K2.6, MiniMax M2.7, GLM-5.1, MiMo V2.5 Pro, Mistral Large 3, Gemma 4 31B, Step 3.7 Flash, FLUX.2 Pro, ElevenLabs, Whisper Large V3
Access: OpenAI-compatible API, Claude Code, OpenCode, Cursor, Kilo Code, VS Code extension, pi.dev, Windsurf, Cline, Roo Code, OpenClaw, Codex CLI +1
Free$0/mo
100 req/5h; 10 RPM; 5 concurrent; DeepSeek-V3.2, Qwen3.5, and more
Lite$10/mo
150 req/5h; 10 RPM; 5 concurrent; Kimi K2.6, MiniMax M2.7, and more; email support
Premium$20/mo
300 req/5h; 20 RPM; 20 concurrent; preview models before lower tiers; high-speed models and priority TPS; priority supp…
Max$50/mo
600 req/5h; 30 RPM; 20 concurrent; first access to preview models; full catalog; dedicated support
 Partial
TrainingPartial
Prompt retentionPartial
PricingModelsPrivacyTerms
60.2 Artificial Analysis Coding Index
MiMo-V2.5-Pro
MiMo-V2.5-Pro
 View
Fabric
HQ: Vancouver, BC, Canada
Models: Fabric Small, Fabric Medium (Qwen 3.6 27B), Fabric Large (Qwen 3.5 397B-A17B, 256K ctx), Fabric XLarge
Access: Fabric IDE, GUI + Terminal + CLI, OpenAI-compatible API, Enterprise air-gapped
Note: Deprecated as a monthly subscription coding plan. Teams $99/user/mo subscription removed; Fabric now sells credits ($1/…
Credit plansfrom $10/mo
$10/mo ($10 credits); $100/mo (+25% → $125); $500/mo (+40% → $700); $1/credit PAYG; $99.99 lifetime license (BYOK only)
 — — — ZDR
TrainingNo
Prompt retentionZero
PricingClawPrivacyTerms
53.7 Artificial Analysis Coding Index
Fabric Medium (Qwen 3.6 27B)
Qwen3.6 27B (Reasoning)
Deprecated; Teams $99/user subscription removed, now credit-backed plans.
 View
Fabric Claw
HQ: Vancouver, BC, Canada
Models: fabric-small (Qwen 3.5 35B-A3B), fabric-medium (Qwen 3.5 27B), fabric-large (Qwen 3.5 397B-A17B, 256K ctx)
Access: OpenAI-compatible API, Fabric IDE, OpenAI SDK clients, agentic workflows, curl/Python/TypeScript/Go
Note: No longer available in Fabric billing; public Claw page remains active for reference.
Medium$19.99/mo
Closed to new purchase; was unlimited fabric-small + fabric-medium; 60 req/min
Large$99/mo
Closed to new purchase; was unlimited all fabric models; 397B / 17B active; 256K ctx
 — — ZDR
TrainingNo
Prompt retentionZero
ClawPricingPrivacyTerms
48.2 Artificial Analysis Coding Index
Fabric Large (Qwen 3.5 397B-A17B)
Qwen3.5 397B A17B (Reasoning)
 View
Fireworks Fire Pass
HQ: Redwood City, CA, US
Models: Kimi K2.5 Turbo
Access: OpenCode, Cline, Kilo Code, OpenClaw, LangChain Deep Agents, OpenAI-compatible tools
Note: Currently closed to new subscriptions; deprecated for new users.
Fire Pass~$30/mo
Official $7/week pass; currently closed to new subscriptions
 — — — ZDR
TrainingNo
Prompt retentionZero
PrivacyZDR docs
46.8 Artificial Analysis Coding Index
Kimi K2.5
Kimi K2.5 (Reasoning)
 View
Alibaba Cloud Model Studio — Coding Plan Lite
HQ: Hangzhou, China
Models: qwen3.5-plus, kimi-k2.5, glm-5, qwen3-coder-plus, glm-4.7
Access: —
Coding Plan LiteCustom
No new purchases since Mar 20 2026; existing users retain use, renewal, and upgrade rights. Pro plan remains active.
 — — — No training
TrainingNo
Prompt retentionLimited
PrivacySecurity
45.3 Artificial Analysis Coding Index
GLM-4.7
GLM-4.7 (Reasoning)
GLM-5 and Qwen3 Coder Plus are listed but do not currently have Artificial Analysis Coding Index scores; GLM-4.7 is the highest-scoring listed benchmarked model.
 View
Cerebras Code
HQ: Sunnyvale, CA, US
Models: GLM 4.7 coding route (high-context)
Access: OpenCode, OpenClaw, OpenAI-compatible tools
Pro$50/mo
Sold out; 24M tok/d hard cap
Max$200/mo
Sold out; 120M tok/d hard cap
 — — ZDR
TrainingNo
Prompt retentionZero
PrivacyTerms
45.3 Artificial Analysis Coding Index
GLM-4.7
GLM-4.7 (Reasoning)
Claimed ~1000 tok/s
 View
AI Foundry
HQ: Auckland, New Zealand
Models: Nemotron Omni 30B (128K ctx), Qwen3 Coder Next (256K ctx)
Access: OpenAI-compatible API, OpenClaw, OpenAI SDK clients, LangChain, LiteLLM, developer tooling
Note: Pivoted to sovereign GPU/infrastructure rentals; all pricing is bespoke per workload.
OpenClaw$7/mo
Flat fee; Nemotron Omni 30B; 128K ctx; multimodal chat, vision, audio, OCR; 8 concurrent requests
Coder Pro$19/mo
Flat fee; Qwen3 Coder Next; 256K ctx; coding agents, refactors, code review; 8 concurrent requests
 — — ZDR
TrainingNo
Prompt retentionZero
InferenceAppFair useDPAPrivacyTerms
36.2 Artificial Analysis Coding Index
Qwen3 Coder Next
Qwen3 Coder Next
Historical: AI Foundry Coder Pro listed Qwen3 Coder Next with 256K context. AI Foundry has since pivoted to sovereign GPU rentals and no longer offers a coding plan.
 View